You are viewing a single comment's thread from:

RE: api.deathwing.me Important Announcement | Anti-Phishing

in #hive3 years ago

I don't think you can help someone who would transfer $20K without triple checking the receiver's address, but it's interesting to see how easily Hive transactions can be censored

Sort:  

you don't need an api node to broadcast things. Even the most basic hived on the P2P network will accept a broadcast. There are many of those if you found yourself in the position of needing to find a friend, and they aren't that hard to start up if you wanted to use blockchain in a trustless fashion. This modification is to Jussi - a middleware that routes requests to a public api between hived (actual blockchain node) and hivemind (layer2 social media node). It isn't even a change to hived

Ok, if any hived node can do the job, then why only a few are used by the apps? I'm curious because I'm observing similar behaviour in the WAX blockchain and can't figure it out

An api node does things like provide posts, voting data, account history, etc. You need these bigger, beefier servers to serve up the information for the dapps and websites, but to broadcast a transaction you only need any hived on the p2p network

I think I got it now, thanks. And your other comments on this post were very helpful to see the picture more clearly

This has nothing to do with censorship lol. Nodes are not obliged to provide their service to you

exactly. I guess better would be a warning to check the account name again else you can lose your funds.

A reply I wrote in the comments:

In this case, for transactions, as I mentioned in the post... Some apps do provide "feedback" whenever potential phishing is about to occur. For example, as asgarth explained, Peakd uses a "similarity check" to see if you are sending it to the right exchange account.

Unfortunately, some apps, scripts, bots etc. Do not have this check. For example, Keychain does not detect whether or not you're sending a transaction to blocktraded instead of blocktrades just because you accidentally pressed "D" (which is right beside S in most QWERTY layout keyboards around the world.) -- This implemented feature helps mitigate that. I'd love to give a warning, but there is no Are you sure you want to perform this transaction? system in the codebase. It's True or False, Yes or No. In my case, I decided on creating a blacklist of phishing accounts to protect accidental transfers.

You're still free to send funds to the wrong recipient if that's what you want to do. Nobody is stopping you.

Some prefer to buckle their seatbelts while others prefer to fly through the windshield. It's a choice.

That would be a fault in the existence of software, not Hive design. If you want to modify the software and run it, then it can do other things. This is true even in Bitcoin

If you are broadcasting to another node, you are trusting someone. This is why you find many blockchain purists that will only interact with their own node. It is the only way to guarantee trustlessness

There is really nothing to talk about. All cryptocurrencies, not just Hive, is subject to the same exact situation. A node operator in all cryptocurrencies can easily mess with the data they're relaying and it would be up to the scripts or the person to figure out if it's right or not.

All users can decide on which node they'd like to use, for whatever reason. Be it speed (low latency, better hardware etc. overall faster experience) or trust (they are certain that the node operator is not doing funky stuff in the background to their data)